Jurnal Ekonomi Malaysia 18(Disember 1988) 27-42

Short-Term Effects of Malaysian Redistribution Policies Based on a Computable General Equilibrium Model

Ragayab Haji Mat Zin

ABSTRACT

This study investigates the base-year solution to the general equilibr­ium model of Malaysian redistribution policies. These policies include the tax-expenditure policies and, in particular, those of the New Eco­nomic Policy. The first-round effects indicate that increasing borrow­ed capital of Bumiputeras appears to be one of the most effective means of moving the economy towards achieving the Bumiputera share capital ownership target, while increasing the l{md ownership share of Bumi­puteras and government expenditure benefits are efficacious for the income redistribution objective. Increasing employers' contributions to the EPF and reducing the corporation tax rate are efficient means of attaining the employment restructuring targets.

ABSTRAK

Kajian ini menyiasat kesan dasar-dasar agihan semula di Malaysia dalam jangkamasa pendek dengan menggunakan model keseimbang­an umum. Dasar-dasar ini termasuklah dasar cukai-perbelanjaan dan, khususnya, yang berkaitan dengan Dasar Ekonomi Baru. Kesan-kesan pusingan pertama menunjukkan bahawa menambahkan modal pinjam­an bagi Bumiputera merupakan satu daripada cara yang paling ber­kesan menggerakkan ekonomi kepada pencapaian sasaran bahagia:n

pemilikan modal saham Bumiputera, semen tara menambahkan bahagi­an pemilikan tanah oleh Bumiputera dan faedah-faedah dari perbelan­jaan kerajaan adalah berkesan untuk sasaran agihan semula pendapat­an. M enambahkan sumbangan majikan kepada KWSP serta menurun-kan kadar cukai syarikat adalah pula cara yang cekap bagi mencapai sasaran menyusun semula gunatenaga.

INTRODUCTION

This stucty examines the base-period solution to the general equilibr­ium model of Malaysian redistribution policies which was presented

28 Jurnal Ekonomi Malaysia 18

in an earlier issue of this journal (Ragayah 1988b). This model was formulated in order to investigate the probable effectiveness of the Malaysian Government's redistributive policies. These policies in­clude the tax-expenditure and, in particular, the policy of redistri­buting capital ownership and the restructuring of employment among Buniiputeras, as well as that of eradicating poverty, which has an in­come redistribution connotation. In constructing the model, attempts

were made to capture the principal featun~s of the economy and the relevant restructuring policies that were being carried out to attain the set targets.

In our earlier paper, we have described that the structural form of the model can be written as

(1)

where Ao, Al are 91 x 91 matrices while Bo is a 91 x 16 matrix, and Y is the column vector of the endogenous variables, Yt - 1 is that of the lagged endogenous variables and X is that of the exogenous vari- ' abies. The symbol Y indicates that all these variables are expressed as proportionate changes, for example, Y = dY/Y. We then solved the system 1 in its reduced form:

~ _ -1 ~ -1 ~

Yt - Ao Al Yt - 1 + Ao BoX t • (2)

The coefficient matrix Ao 1 Bo in equation (2) is the matrix of the estimated impact multipliers. An impact multiplier, aYj(t)/ax/t), measures the immediate response df an endogeneous variable con­sequent to a change in an exogenous variable, keeping all other exo­genous variables constant. Owing to the method of taking total deri­vatives and proportionate changes, the multipliers yielded are pro­portionate changes of the base period values, that is the coefficients of Ao 1 Bo measure the proportionate changes in the endogenous variables on the left hand side consequent to one percent change in the exogenous variables on the right hand side.

As mentioned previously (Ragayah 1988a, 1988b), the lineai'i­zation process requires the absolute values of all the variables in­volved. Sahota (1975: 98) points out that the choice of which year values depends on the objective of the study. If the purpose is to make predictions, the final-year values of the variables in current prices should be used, but if it is to explain the data the appropriate values are the sample means in constant prices. However, in our case, these two options are not opened to us due to the paucity of data. Hence, our choice was made on the basis of which year has the most data

Effects of Malaysian Redistribution Policies 29

available, and by this criterion 1973 was chosen as the base year. Fo'r those variables that do not have their 1973 values, we have used the Consumer Price Index to inflate or deflate the available values to their 1973 equivalents.

The construction of this 1973 data set is described in Ragayah (l988a, Appendix B), together with extraneous econometric estimates on various elasticities. Among the data sources used are: national accounts, an input-output table, census of manufacturing industries, annual reports of the Department of Inland Revenue, Treasury Economic Report, Quarterly Survey of Employment in Manufactur­ing Industries, 1973 Household Expenditure Survey, Report of the Financial Survey of Limited Companies, Monthly Statistical Bull­etin of the Statistics Department, Bank Negara Malaysia Quarterly Economic Bulletin, 1981 Annual Report of the Permodalan Nasio­nal Berhad and the various Malaysia Five-year Plan.

Goldberger (1959), Sahota (1975), and Sahota and Rocca (1980) pointed out several caveats that must be borne in mind when inter­preting these impact multipliers. Following Sahota's latter work, we enumerate these caveats here. Firstly, their values are likely to be different from the total multipliers since they only register the first year responses and neglect the lagged influences. Secondly, these co­efficient are expressed in terms of impact elasticities instead of mar­ginal effects. As such, it is thus quite possible for the values of some of these multipliers to exceed unity. Thirdly, some of multipliers have signs that appear pri'ma facie contrary to expectation. However, it must be remembered that the ultimate impact effect is the outcome of several supply and demand markets, and the price effects within and across the different sectors. Given such a situation, it will be some­times impossible to distinguish a single force which can account for such contrary results. Fourthly, the difficulty in analyzing these sup­posedly suspect results is further compounded by the fact that these computed coefficients are also subject to stochastic errors. Since it is not possible to caiculate the standard errors, then some of these unexpected outcomes could be merely due to cumulated errors. Finally, as mentioned elsewhere, we abstract from the foreign sector in order to keep the size of the model at a manageable level.

THE SOLUTIONS

The impact elasticities of the base period solution are given in Table 1. In order to avoid the problem of not being able to see the trees for

TABLE 1. Impact-elasticities

Indep. Var. k~~ 1

Dep. Var. t~

kBII 0.04946 0.01152 -0.04573 0.00005 - 0.00006 - 0.26709 kNB 0.00000 0.00057 0.08657 - 0.00001 0.00008 - 0.24101 KB 0.04762 0.01054 -0.03932 0.00004 -0.00013 -0.02511 Y 1B -0.00001 0.24225 0.72753 -0.00117 -0.00203 0.49184 Y lNB - 0.00001 0.01107 0.93786 - 0.00052 - 0.00068 0.13471 Y mB 0.00549 0.00178 0.94722 - 0.00020 0.00215 0.91569 YmNB -0.00003 0.00128 0.94789 -0.00012 0.00189 0.81159 Y 6B 0.00050 0.01432 1.30320 0.03013 0.02201 1.15600 Y 6NB 0.00006 0.00348 0.90668 0.00365 0.00471 0.87272 Y B 0.00039 0.05215 0.23517 0.00004 - 0.00006 0.18101 YNB -0.00001 0.00330 0.40021 -0.00013 0.00012 0.15684 YR -0.00001 0.03924 0.35740 -0.00031 -0.00048 0.10940 Yu 0.00045 0.00064 0.30909 0.00'037 0.00091 0.27512 L1B 0.00023 0.00232 0.15264 0.01268 0.00487 0.11471 L1NB 0.00018 0.00024 0.09694 0.00924 0.00176 0.07642 L2B -0.00070 -0.01139 -0.47488 -0.05148 -0.02358 -0.35]64 L2NB -0.00019 -0.00309 -0.12889 -0.01398 -0.00640 -0.09544 L3B - 0.00061 - 0.01978 - 0.71630 - 0.02839 - 0.02706 - 0.50299 L3NB -0.00035 -0.01145 -0.41464 -0.01643 -0.01566 -0.29116 L6B 0.00129 0.03209 1.21640 0.07669 0.05087 0.86997 L6NB 0.00113 0.03035 1.13460 0.06354 0.04607 0.80708

Indep. Var. Dep. Var.

u' U" i* G

kBII 0.34226 0.01915 0.00337 0.00.004 0.00077 0.00428 kNB 0.37000 0.03761 0.00003 0.00497 0.00000 0.00909 KB - 0.02671 - 0.01777 0.00322 - 0.00475 - 0.00074 0.01287 Y lB - 0.05525 0.02892 - 0.00745 0.00057 - 0.00039 0.17449 Y 1NB 0.04450 0.01693 0.00009 -0.00285 0.00001 0.02916 Y mB 0.70877 0.00045 - 0.03984 - 0.00040 - 0.00409 - 0.16653 YmNB 0.78091 0.11491 0.00014 -0.04456 0.00002 -0.15949 Y 6B - 1.28730 - 0.23838 - 0.00422 - 0.00838 - 0.00039 0.24472 Y6NB 0.34738 -0.04788 -0.00051 -0.00285 -0.00005 -0.05722 Y B 0.02451 0.00373 - 0.00442 0.00001 - 0.00037 0.02823 YNB 0.12219 0.01989 0.00004 -0.00689 0.00000 -0.01413 Y R 0.00350 0.00882 - 0.00109 - 0.00067 - 0.00006 0.03388 Yu 0.21470 0.02097 -0.00326 -0.00970 -0.00033 -0.04561

TABLE 1. (Continued)

LIB - 1.80990 - 0.04584 - 0.00171 - 0.00257 - 0.00017 0.43638 LINB -2.05510 -0.00596 -0.00134 ....:0.00170 -0.00014 0.52813 L2B 2.03260 0.33430 0.00537 0.00707 0.00054 - 0.26154 L2NB 0.55172 0.09076 0.00146 0.00192 0.00015 -0.07096 L3B 2.67560 0.22446 0.00497 0.01254 0.00048 - 0.57373 L3NB 1.54880 0.12991 0.00288 0.00726 0.00028 - 0.33212 L6B -4.76910 -0.54294 -0.01025 -0.02022 -0.00101 0.87062 L6NB - 4.38540 - 0.46103 - 0.00905 - 0.01916 - 0.00087 0.84124

Notes:

1. Definitions of endogenous variables:

kB" : Capital owned by Bumiputera household kNB : Capital owned by non-Bumiputera household KB : Share of Bumiputera capital ownership y 1B : Income of Bumiputera household in the rural sector y 1NB Income of non-Bumiputera household in the rural sector Y mB : Income of Bumiputera household in the modern sector Y mNll : Income of non-Bumiputera household in the modern sector Y 6B : Income of Bumiputera household in the urban informal sector Y 6NB : Income of non-Bumiputera household in the urban informal sector Y B Income of Bumiputera household Y NB Income of non-Bumiputera household Y R Income of rural household Y u Income of urban household L1B Bumiputera labor, supply to the rural sector L1NB Non-Bumiputera labor supply to the rural sector L2B Bumiputera labor supply to the secondary sector L2NB : Non-Bumiputera labor supply to the secondary sector L3B : Bumiputera labor supply to the tertiary sector L3NB : Non-Bumiputera labor supply to the tertiary sector L6B : Bumiputera labor supply to the urban informal sector L6NB : Non-Bumiputera labor supply to the urban informal sector.

2. Definitions of exogenous variables:

k~~ 1 : Borrowed capital of Bumiputera household !B Share of Bumiputera land ownership ty ty = 1 - t~; t~ is personal income tax rate t~m Domestic or commodity tax rate t~ Import tax rate tw1 tw1 = 1-t~1; t~l is employees' contribution rate to the EPF

tw2 tw2 = 1 + t~2; t~2 is employer's contribution rate to the EPF

tc tc = 1 - t~; t~ is corporation tax as a percentage of profits u' : Undistributed profits rate of Bumiputera household u" : Undistributed profits rate of non-Bumiputera household i* Cost of borrowing capital to Bumiputera household G : Government expenditure benefits to each household.

32 Jurnal Ekonomi Malaysia 18

the forest, and not to discourage the reader from examining these results, we have presented only the impact multipliers of the twenty­one dependent variables in which we are particularly interested in the capital ownership, income and employment (this is proxied by labor supply) variables, consequent to changes in the twelve policy variables. Let us now look at the effects of these policy instruments.

Borrowed capital of Bumiputeras, k:' _ 1 It is comforting to note that the effects of borrowed capital of Bumiputeras mostly display the anticipated results. An increase in this item will definitely en­hance the value of Bumiputera capital accumulation in the first year. Non-Bumiputera capital ownership contracts marginally, which is expected due to the crowding out effect. This expansion in the value of Bumiputera capital accumulation leads to an increase in the capital ownership share of Bumiputeras, as the impact multiplier for the capital share of Bumiputeras is positive.

In terms of income, with the exception of Bumiputeras in the rural and non-Bumiputeras in the rural and modern urban sectors, borrowed capital of Bumiputeras benefit everyone else, with Bumi­puteras in the urban modern sector making the greatest gain. How­ever, these negative effects are very small and can be considered neu­tral. The outcome of Bumiputeras' income in the urban modem sector is an expected one since this group holds the largest amount of bor­rowed capital, and has the highest capacity of increasing its holdings. This group also accounts for the fact that Bumiputera and urban in­comes increase at faster rates than non-Bumiputera and rural in­comes respectively as a consequence of an increase of this policy in­strument.

The effects of borrowed capital of Bumiputeras on labor supplies seem to be mixed. Labor supplies of both races rise in the rural and urban informal sector, but drop slightly in the urban modern sector. We can try to explain this outcome in the following way. Borrowed capital, which becomes part of total capital, is mainly being ploughed into the more capital intensive modern sector. Here, capital appears to be a substitute for labor inasmuch as jobs in this sector are more e~sily mechanized. On the other hand, expansion in capital invest­ment would lead to an increase in employment of these displaced workers in the rural and urban informal sectors since these are low­skilled or unskilled workers who are utilized in labor-intensive jobs. Hence, here capital is complementary to labor. Thus, even though this policy instrument does enable the Bumiputeras to increase their

Effects of Malaysian Redistribution Policies 33

capital ownership, it has a negative effect on the employment res­tructuring policy.

Share of Bumiputera land ownership, r B As expected, the share of Bumiputera land ownership has positive effects on the value of capital accumulation of Bumiputeras .. Similarly, it has the anticipated favorable outcome on the income levels of all Bumiputera groups, with those in the agriculture sector making the largest gain. Increases in land ownership means direct increase in income through its return, and a rise in income would lead to greater capital accumulation. Naturally, Bumiputera and rural households make the greatest gains, which is accounted for by the fact that all returns from land is at­tributed to the rural households.

The results also indicate that the effects of this instrument on the value of capital accumulation of non-Bumiputeras and their income levels are also positive. However, since the value of the multiplier for the capital ownership of Bumiputeras is greater than the elasticity of capital ownership of non-Bumiputeras, its effects is to increase the capital share of Bumiputeras. These positive results on the capital accumulation and income levels of non-Bumiputeras may be un­expected at first. Nevertheless, if we recall that the supply of culti­vable land is increasing annually, then these results might not be as puzzling as it first appear. Since Bumiputeras also own a much small­er proportion of cultivable land than non-Bumiputeras, a rise in their share would not jeopardize the share of non-Bumiputera land in­come much. This leads to the positive values of their capital owner­ship and income multipli~rs, though the magnitudes of these are smaller than those for Bumiputeras, which enforces our argument.

The expansion in Bumiputera land ownership has negative ef­fects on labor supplies to the modern sector and positive effects to the rest. In the rural sector, a bigger land holding requires more labor or can accomm·odate more labor to work on it resulting in an ex­pansion in Bumiputera rural employment. While non-Bumiputera labor has relatively less employment opportunity in the agriculture sector consequent to the expansion in the share of Bumiputera land holdings, this policy still results in a positive, though slight, change in non-Bumiputera labor supply in this sector. As pointed out above, rural labor supplies of both groups can expand because the supply of cultivable land is rising every year. Consequently, there are nega­tive effects on labor supplies to the urban modern sector. These vacancies cannot be filled by those from the urban informal sector

34 Jurnal Ekonomi Malaysia 18

because, according to Mazumdar's findings (1981: 223), unlike the stereotype of rural-urban migrants who flocked to the urban informal sector, recent Malaysian rural-urban migrants apparently obtained jobs in disproportionate numbers in large enterprises with employ­ment of one hundred or more. Furthermore, Mazumdar also found that the migrants appeared to be generally better educated than the natives in the urban informal sector, and as such, it may not be pos­sible for the latter to fill in the vacuum in the modern urban sector.

Personal income tax, t~ In interpreting these multipliers, it must be kept in mind that personal income tax is defined as ty = (1 - 'ty), which implies that a percentage increase in ty means a reduction in the tax rate, ty, not a rise. Hence, the effects of personal income tax on the three groups of dependent variables mostly conform to pre­dictions. A decrease in personal income tax rate, and thus a rise in disposable personal income, has favorable effects on capital owner­ship of both racial groups, with the multiplier for the non-Bumi­puteras being slightly higher than that for the Bumiputeras. This then accounts for the negative growth in the share of capital owner­ship of Bumiputeras. The positive outcomes of this change on the various income levels are self-explanatory. Generally, while every­one's income increases, non-Bumiputera households obtain the greatest gain with this policy.

With regards to labor supplies, we expect a smaller tax rate to encourage more labor to enter the job market since this means higher realized disposable income. This is because labor supplies depend on net-of-tax wage rates, while demand for labor depends on gross-of­tax wage rates, and wages form part of income. Our results show that when the personal income tax rate is lowered, labor supplies of both racial groups to the rural and urban informal sectors increase while that to the urban modern sector decrease. This result can be ex­plained in the following way. Reducing personal income tax leads to an increase in incomes and savings, and thus capital accumulation. This more capital intensive investment substitutes for labor in the modern sector. Less labor will be employed (only the'relatively skilled ones) in this sector. Others will flow back to the rural and urban in­formal sectors, thus accounting for the positive multipliers of the labor supply variables in these sectors.

Commodity taxes, t~m Commodity taxes, which include export duties, excise duties and sales taxes, were aggregated and denoted as

Effects of Malaysian Redistribution Policies 35

domestic taxes since the appropriate data that were available come in this form. The impact on the capital ownership of Bumiputeras is positive while that on non-Bumiputeras is negative, though both effects are marginal, the values of the multipliers being almost zero. These results cause the share of Bumiputera capital ownership to improve by a tiny margin. The Bumiputera result comes about due to the fact that the positivity of their urban informal sector house­holds' income and·saving are able to overcome the negativity of these two variables of the rest of the Bumiputeras. In the non-Bumiputera case, the negative impacts on incomes and savings of their rural and urban modern households dominate over the favorable outcomes on these variables of the urban informal sector households.

The unfavorable effect on income is true for all households except those in the urban informal sector. These adverse impacts come about as a consequent of a decrease in profitability due to the imposition of these taxes. The incomes of the urban informal house­holds are not negatively affected because very little of these taxes are levied from this sector. Furthermore, the taxes collected are being re~istributed as government expenditure benefits which could over­come the duties levied in these two sectors. All these effects tend to neutralize each other, resulting in very small values in the multipliers.

The labor supply variables also reacted in a mixed way, with in­creases in the rural and urban informal sector groups and reductions in the urban modern .,sector groups. However, these responses are compatible since commodity taxes are growth-retarding, as indicated by the negative multipliers of output in most of the modern sector. Hence, employment intake in the urban modern sector would have damped and in turn, this would discourage entry into this sector. Consequently, labor supplies in the other two sectors would expand.

Import taxes, t'ro Import duties have a slight favorable effect on non-Bumiputera and a marginal discouraging outcome on Bumi­putera capital accumulation, thus resulting in a small decrease in Bumiputera share capital ownership. These effects on capital accu­mulation arises from the positive reactions of incomes and savings of households in the urban sectors.

Incomes of all households in the rural areas respond negatively to the imposition of import taxes while incomes of the rest of the households react favorably. This reduction in rural incomes is caused by the fall in the rural wage rate, which in turn is due to an expansion in labor supply in this sector. On the other hand, import duties raise

36 Jurnal Ekonomi Malaysia 18

the income levels of the rest because wages increase in these sectors owing to the contraction in labor supplies; which in turn is caused by the decrease in outputs and net prices. However, the reduction in labor supply exceeds the decrease in outputs and net prices, thus re­sulting in the increase in· the wage rates of these households.

Social security tax contributions, t'wH t'w2 In Malaysia, both em­ployees and employers contribute to the EPF, the employees con­tributing at the rate t'wl and the employers at t'w2. In our calculations we have defined twl = (1 - t'wl) implying that, as in the case of per­sonal income tax, twl increases w~th the reduction in the employees' contribution rate. This means that the net-of-contribution wages of employees increase with the fall in these rates, as wage incomes are defined to be net of these contributions. Hence, we expect income levels of all groups to react positively to this instrument, and our results bear this out. Bumiputeras gain relatively more since the value of the multiplier of their income is a little higher than that of non­Bumiputeras.

With higher incomes, we expect that capital accumulation would be increasing. However, this is not so for the multipliers of both capital ownership variables are negative consequent to the implement­ation of this policy. This is explained by the fact that a decrease in t'wl means a reduction in forced savings in its entirety, while only part of the increase in income emanating from this policy would be saved. Thus, the net result would be a decrease in savings and capital ac­cumulation. The negative multipliers of all saving groups testify to this prediction.

A reduction in this tax causes labor in the modern sector to reduce their supplies which naturally flow back into the other two sectors. These results arise from the following reactions. Lowering this tax causes a reduction in savings and hence capital accumulation. This leads to a contraction of most of modern sector outputs, and con­sequently, their labor requirements, especially the relatively less skilled ones.

Regarding t'w2' where we define tw2 = (1 + t'w2)' employers con­sider their share of the contributions as labor cost, and therefore increasing the price oflabor relative to the price of capital .. Thus, the results of the capital accumulation variables conform to the theory that a fall in the relative price of capital would lead to more capital investment. This is confirmed by the negative value of the price of capital elasticity and positive multiplier for the rate of return to

.:

Effects of Malaysian Redistribution Policies 37

capital. At the same time, this tax, while being a cost to the employers, represents an increase in savings and postponed income to the em­ployees. This also enhances capital accumulation and raise the in­come of workers. These expected results are borne out by the incomes of all non-Bumiputeras and those households in the modern sector. On the whole, incomes of both Bumiputeras and non-Bumiputeras change favorably. Moreover, the share of Bumiputera capital owner­ship decreases since the value of Bumiputera capital accumulation elasticity' is lower than that for n~n-Bumiputeras.

Raising t~2 has the effect of increasing the migration of workers into the modern sector and contracting the supplies to the rural and urban informal sectors. The contraction of the rural labor supplies arises from the fact that output in this sector has declined,' resulting in a negative change in the return from land. On the other hand, out­put is increasing in the modern sector and wage here is rising propor­tionately more than in the other two sectors. The latter reason also explains the exodus of labor supplies from the urban informal sector into the modern sector.

Corporation tax, t~ Like the personal income tax, this tax is defined as tc = (l - t~), which means that an increase in tc is a decrease in the tax rate (. Here, the impacts on capital ownership are expected, that is, a lowering of the corporation tax rate would lead to an ex­pansion in capital ownership of both Bumiputeras and non-Bumi­puteras. In terms of the share of capital ownership, the impact multi­plier of Bumiputera capital ownership increases slightly, which is not suprising since the corresponding value of the Bumiputera capital accumulation is higher than that of non-Bumiputeras.

Incomes of all households in the rural and the uroan modern sectors react positively to the reduction of the corporation tax. This is anticipated since incomes increase both directly via an rise in dividends as well as through the return to additional capital accumu­lation encouraged by the reduction in. this tax. Urban informal households suffer a reduction in income following the lowering of this tax. This is not a suprising outcome since these households are only slightly influenced by this tax. At the same time, the impact on wage in this sector is negative, resulting in a negative response in in­come. However, these negative multipliers do not affect the positivity of the multipliers for the Bumiputera and non-Bumiputera incomes, as well as rural and urban incomes as a whole.

Where labor supplies are concerned, reducing this tax has posi­tive impact on labor supplies to the urban modern sector and adverse

38 Jurnal Ekonomi Malaysia 18

effects on those in the urban informal and rural areas. This result can be accounted for by the fact that the reduced corporation tax in­fluence modern sector income to rise, thereby attracting labor from the other two sectors. Furthermore, outputs in the agriculture and urban informal sectors contract, thus diminishing employment op­portunities in these sectors. This outcome is consistent with the employment restructuring objective of the NEP since this instrument has induced relatively more Bumiputera labor into the modern sector.

Undistributed profits' shares, u'u" Available data have indicated that the undistributed share of profits differ for Bumiputeras and non-Bumiputeras. An increase in undistributed profits of Bumi­puteras, u /, is consistent with increasing their capital ownership since the higher is this share, the higher is the rate of their savings. It has a slight positive effect on the value of capital accumulation of non­Bumiputeras. Therefore, it is not suprising that the elasticity of the share capital ownership of Bumiputeras is p~sitive. Similarly, un­distributed profits of non-Bumiputeras, u", encourages capital ac­cumulation among non-Bumiputeras, and has a positive but almost zero effect on Bumiputera capital ownership. Naturally, the share of Bumiputera capital ownership is decreasing under this regime.

Generally, the effects of undistributed profits of both racial groups on incomes variables also correspond to our expectations, that is, higher shares of undistributed profits mean lower realized disposable income, and hence their negative signs. In both cases, the modern sector households suffered the largest reductions in their incomes, which is not suprising since they derive the highest incomes from capital ownership among all the income groups. It follows then that when u' is increased, it has negative consequence on Bumiputera and urban incomes as a'whole. Likewise, non-Bumiputera and urban incomes reacted unfavorably when u" is raised. The adverse effect of u' on Bumiputeras income is worse than that for non-Bumiputeras income, and vice versa" while they both curtail urban income more than rural income.

Raising each of the undistributed profits shares would attract greater supplies of labor to the modern sector. This is not suprising since greater investment activity also calls upon greater labor ab­sorption. Furthermore, although u' has slight positive effects on wages, the value of the multiplier is largest for the modern sector wage. Moreover, both induce a faster increase in Bumiputera em­ployment in the modern sector than non-Bumiputera employment.

Effects of Malaysian Redistribution Policies 39

This higher demand for labor in this sector draws its supply from the other two sectors, thereby decreasing their supplies, as indicated by the negative coefficients of these variables. In general, the results of increasing u' are consistent with the capital ownership as well as the employment restructuring objectives of the NEP.

Cost of borrowing capital to Bumiputeras, i* The results bear out the forecast that capital ownership of Bumiputeras is adversely af­fected by an increase in the cost of borrowing capital. The value of non-Bumiputera capital ownership rises marginally. Both of these reactions result in the share of Bumiputera capital ownership to shrink.

As expected, incomes of Bumiputera households in all sectors are unfavorably affected by this policy. Non-Bumiputera households in the rural and urban modern sectors react positively while those in the urban informal sector have a negative response to this policy. However, all the non-Bumiputera elasticities are marginal and could be considered neutral. When aggregated into the racial and rural­urban groupings all incomes are reduced owing to this higher cost of borrowing faced by Bumiputeras, and the reduction being larger, understandably, for Bumiputera income than non-Bumiputera in­come and urban income than rural income.

The cost of borrowing capital to Bumiputeras appears to have the effect of swelling the labor supply in the urban modern sector, and lessening the labdr supplies in the other two sectors. We can ex­plain this phenomena in the following way. When there is an increase in i*, this implies that the Government subsidy element is reduced, thus discouraging Bumiputeras from borrowing and investing in the stock market. It follows that the supply of capital contracts ana pro­ducers substitute labor for capital, which raises wages in all sectors marginally. This factor substitution seems to induce Bumiputera labor to enter the modern sector more than it does non-Bumiputera labor, thereby steering Bumiputera employment at least in the di­rection of the NEP employment restructuring target.

Government expenditure benefits, G Government expenditure benefits include expenditures on education, medical and agricul­tural services, and pensions. The estimated impact elasticities indicate that Government expenditure benefits raise the value of capital ownership of Bumiputeras, but lower that of non-Bumiputeras. The former complies with our prediction since Government expenditure benefits represent an increase in income, at least in the year they are

40 Jurnal Ekonomi Malaysia 18

received, and households save part of this increment and augment capital accumulation. Non-Bumiputeras do not enhance the value of their capital ownership because most of this increase in income is spent on consumption, as most of the impact multipliers for the various consumption groups of non-Bumiputeras are positive. Fur­thermore, Bumiputeras, especially those in the rural sector, receive a greater amount of Government expenditure benefits than non­Bumiputeras. It is not surprising that the multiplier of the share of Bumiputera capital ownership is positive under these circumstances.

This policy instrument is beneficial to Bumiputera households in the rural and urban informal sector as well as non-Bumiputeras in the rural areas by raising their income levels. However, it is disad­vantageous to all households in the urban modern sector and non­Bumiputera households in the urban informal sector. This is an un­anticipated outcome since Government expenditure benefits rep­resent a direct increase in income levels. The ohly way for these in­come levels to contract is for the gain in income via Government expenditure to be less than the loss in income through the other two components, namely wages and the return to capital. For the rural households, they also receive incomes in the form of the return from land whose multiplier is positive to overcome this negativity, hence making their income elasticities positive. Wages of these groups did fall substantially. In the modern sector, this is caused by the fall in output. Similarly, the impact multiplier of the return to capital is also negative. These factors then explain why the income levels of the households in the modern and non-Bumiputera households in the urban informal sectors decrease instead of increasing with Govern­ment -expenditure benefits. This response is more acceptable when we realize that the amount of Government expenditure benefits to these households are much less than those to the Bumiputera house­holds in the agriculture sector. Hence, it is not suprising that when the incomes of Bumiputera and non-Bumiputera, and rural and urban, households are contrasted this policy instrument is found to be pro-Bumiputera and pro-rural, and therefore redistributive. Much of the gain is made at the expense of the rich urban modern sector. .

The consequence of Government expenditure on labor supplies does not come as a surprise. Since Government expenditure is ad- -( vantageous to rural households as well as Bumiputera households in the urban informal sector, it acts as a deterrent to the flow of migrant into the urban areas. Negative effect on modern sector wage

Effects of Malaysian Redistribution Policies 41

discourage workers from migrating into the modern sector. All the surplus labor flow back to the other two sectors which act as depots for the unemployed.

CONCLUSIONS

On the basis of these results, a number of policy instruments avail­able to the Government do encourage capital accumulation by Bumi­puteras. Increasing borrowed capital of Bumiputeras appears to be one of the most promising means of achieving the target of Bumi-· putera share capital ownership. This is explained by the fact that an increase in this borrowing has the impact of increasing capital owner­ship of Bumiputeras by the whole amount since all the money has to be used for purchasing capital shares. Other effective instruments in steering the country towards this goal are the provision of Govern­ment expenditure benefits and raising the share of land ownership of Bumiputeras, both of which are understandable.

The most discussed income distributions in the Malaysian economy are those between the ethnic groups and the rural and urban areas. With regards to the former, our results indicate that a majority of the instruments does raise the income of Bumiputeras. However, the most efficient policy to achieve this objective is the re­duction in the personal income tax followed by a lowering of the social security tax contributions by employees. Nevertheless, these are not redistributive since the former is more pro-non-Buniiputeras, while the second is only slightly pro-Bumiputeras. Increasing the share of land ownership of Bumiputeras and Government expen­diture benefits are the most effective for this income redistribution goal. These are also the policies to be pursued in order to redistribute income from the urban to the rural areas, with increasing Govern­ment expenditure benefits to be the most effective.

Where employment restructuring is concerned, the policies which encourage positive increase in labor supplies to the modern sectors are increasing the employers' contributions to the EPF, re­ducing the corporation tax rate, increasing undistributed profits of both racial groups and the cost of borrowing capital to Bumiputeras. All of these policies favor Bumiputera labor relative to non­Bumiputera labor, the most effective being employers' contributions to the EPF.

Still, all these results only record the first year responses. We have to examine the dynamic multipliers to see how effective these policies are in moving the economy towards the N~P targets.

42 Jurnal Ekonomi Malaysia 18

ACKNOWLEDGEMENT

This article is taken from most of Chapter IV of the author's Ph. D. dissertation (1988). For their guidance and encouragement of this work, I am deeply indebted to Professor Gian S. Sahota, chairman of my dissertation committee, and to Professors Malcolm Getz, Andrea Maneschi, Charles M. Becker and Donald M. Hancock, members of this committee. My utmost gratitude also goes to Pro­fessor ' Cliff J. Huang who wrote the program for calculating the

. multipliers. Of course, I alone am responsible for any remaining errors or shortcomings.

NOTES

I The existence of a solution requires that Ao is a non-singular matrix.

REFERENCES

Goldberger, A. S. 1959. Impact Multipliers and Dynamic Properties of the Klein-Goldberger Model. North-Holland.

Mazumdar, D. 1981. The Urban Labor Market and Income Distribution: A Study of Malaysia. A World Bank Research Publication, Oxford University Press.

Ragayah Haji Mat Zin. 1988a. A General Equilibrium Approach to Fiscal Incidence and Redistribution Policies in West Malaysia. Ph. D., Dis­sertation, Department of Economics, Vanderbilt .University.

Sahota, G.S. 1975. Brazilian Economic Policy, An Optimal Control Theory Analysis. Praeger Publishers.

Sahota,G.S., & Rocca, C. A. 1980. Income Distribution: Theory, Modell­ing and Case Study of Brazil. Department of Economics, Vanderbilt University. Mimeograph.

Fakulti Ekonomi Universiti Kebangsaan Malaysia 43600 UKM Bangi Selangor D.E., Malaysia